Transient receptor potential (TRP) channels: a clinical perspective

Abstract

Transient receptor potential (TRP) channels are important mediators of sensory signals with marked effects on cellular functions and signalling pathways. Indeed, mutations in genes encoding TRP channels are the cause of several inherited diseases in humans (the so-called ‘TRP channelopathies’) that affect the cardiovascular, renal, skeletal and nervous systems. TRP channels are also promising targets for drug discovery. The initial focus of research was on TRP channels that are expressed on nociceptive neurons. Indeed, a number of potent, small-molecule TRPV1, TRPV3 and TRPA1 antagonists have already entered clinical trials as novel analgesic agents. There has been a recent upsurge in the amount of work that expands TRP channel drug discovery efforts into new disease areas such as asthma, cancer, anxiety, cardiac hypertrophy, as well as obesity and metabolic disorders. A better understanding of TRP channel functions in health and disease should lead to the discovery of first-in-class drugs for these intractable diseases. With this review, we hope to capture the current state of this rapidly expanding and changing field.

LINKED ARTICLES

This article is part of a themed section on the pharmacology of TRP channels. To view the other articles in this section visit http://dx.doi.org/10.1111/bph.2014.171.issue-10

Introduction

Regulated transport of ions via ion channels underpins a number of fundamental physiological functions (Bagal et al., 2013). Conversely, inherited (‘channelopathy’) or acquired dysfunction of these channels disrupts physiological processes, leading to a broad array of disorders (Bagal et al., 2013). Ion channels are important targets for many currently prescribed drugs, second only to GPCRs (Clare, 2010). Indeed, the worldwide sales of ion channel drugs are estimated to be in excess of $12 billion annually. Although ion channels have been successful drug targets, achieving subtype-selectivity has been a major challenge, particularly with voltage-gated sodium and calcium channels (Clare, 2010).

Recently, a number of novel, ‘druggable’ ion channels have been identified. Of these newly discovered channels, those of the transient receptor potential family [TRP; Figure 1 (Nilius and Owsianik, 2011); channel and receptor nomenclature follows Alexander et al., 2013] are arguably the most appealing therapeutic targets (see Moran et al., 2011; Fernandes et al., 2012; Vay et al., 2012; Kaneko and Szallasi, 2013). Generally speaking, TRP channels are cellular sensors involved in nociception (Patapoutian et al., 2009), taste perception (Nilius and Appendino, 2013), thermosensation (Tominaga, 2007), mechano-and osmolarity sensing (Pedersen and Nilius, 2007; Guilak et al., 2010). TRP channels also play a crucial role in normal physiological processes such as signal transmission (Minke, 2010; Wu et al., 2010). Dysfunction of TRP channels has been implicated in various disease states (summarized in Figure 2) ranging from chronic pain and overactive bladder (TRPV1) through obesity (TRPV4 and TRPM5), diabetes (TRPV1, TRPM4), chronic cough (TRPA1, TRPV1), and chronic obstructive pulmonary disease (COPD; TRPV4) to cardiac hypertrophy (TRPC6), familial Alzheimer's disease (TRPM7), dermatological disorders (TRPV3 in Olmsted syndrome) and cancer (TRPC6, TRPV2 and TRPM8). Gain-of-function mutations in genes encoding TRP channels have been linked to human diseases as exemplified by familial episodic pain syndrome (TRPA1; Kremeyer et al., 2010). Taming these hyperactive TRP channels by antagonists may prove clinically beneficial. Loss-of-function mutations (e.g. loss of TRPML1 function in type-IV mucolipidosis) are also pathogenic, but their correction is more problematic (Dong et al., 2008).

Figure 1.

Simplified topographical structure of TRP channels (A). Please note the similarities and differences between TRP channel subfamilies (B). Reprinted, with permission, from Nilius and Owsianik (2011).

Figure 2.

Schematic illustration of the tissue-distribution of TRP channels and their putative roles in the pathogenesis of human disease.

TRP channels are also primary targets for a number of natural products with therapeutic potential (summarized in Table 1). For instance, TRPV1 is highly expressed in a distinct population of sensory neurons where it mediates excitation and subsequent desensitization to capsaicin (see Figure 3 for structure) and its ultrapotent analogue, resiniferatoxin (Szallasi and Blumberg, 1999). At present, resiniferatoxin (Figure 3) is undergoing clinical trials (NCT00804154) as a ‘molecular scalpel’ to achieve permanent analgesia in patients with intractable cancer pain (Iadarola and Mannes, 2011; Iadarola and Gonnella, 2013). TRPM8 is the ‘menthol receptor’ (Knowlton and McKemy, 2011) and TRPC6 is believed to mediate the mood-improving effect of hyperforin, the main ingredient in St. John's Wort (Leuner et al., 2007).

Table 1.

Endogenous and exogenous ligands of TRP channels

TRP channels	Ligands	References
TRPV1	Endogenous agonists:
	Anandamide*2009	Zygmunt et al. (1999)
	N-arachidonoyldopamine	Huang et al. (2002)
	N-oleoyldopamine	Chu et al. (2003)
	12-and 15-hydroperoxyeicosatetraenoic acid, 5-and 15-hydroxyeicosatetraenoic acid, Leukotriene B4	Hwang et al. (2000)
	9-and 13-hydroxy-octadecadienoic acid(ODE), 9 and 13-oxoODE	Patwardhan et al. (2009)
	Oleoylethanolamide	Ahern (2003)
	Palmitoylethanolamide*2009	Ambrosino et al. (2013)
	Lysophosphatidic acid	Nieto-Posadas et al. (2011b)
	Endogenous antagonists:
	Resolvin D2	Park et al. (2011a)
	Exogenous agonists
	2-Aminoethoxydiphenyl borate (2-APB)	Hu et al. (2004)
	Ornithoctonus huwena toxin [‘double-knot’ toxin (DkTx)]	Bohlen et al. (2010)
	Capsaicin*2009	Caterina et al. (1997)
	Piperine	McNamara et al. (2005)
	Resiniferatoxin*2009	Szallasi and Blumberg (1989)
	Gingerol	Liu et al. (2000)
	Evodiamine	Pearce et al. (2004)
	Cannabidiol*2009	Bisogno et al. (2001)
	Cannabigerol	De Petrocellis et al. (2011)
	Polygodial	Andrè et al. (2006)
	Vanillotoxin	Siemens et al. (2006)
	Exogenous antagonists
	Capsazepine	Dickenson and Dray (1991)
	Iodo-resiniferatoxin	Seabrook et al. (2002)
	BCTC	Valenzano et al. (2003)
	Thapsigargin	Toth et al. (2002)
	Yohimbine	Dessaint et al. (2004)
	AG489, AG505	Kitaguchi and Swartz (2005)
	ABT-102*2009, AMG-517*2009, AZD-1386*2009, DWP-05195, GRC-6211*2009, JTS-653*2009, MK-2295, PHE377, SB-705498*2009	Moran et al. (2011); Brederson et al. (2013)
TRPV2	Endogenous agonists:
	Lysophosphatidylcholine, Lysophosphatidylinositol	Monet et al. (2009)
	Exogenous agonists
	Cannabidiol*2009, Δ9-tetrahydrocannabinol cannabinol*2009	Qin et al. (2008)
	2-APB	Hu et al. (2004)
	Probenecid	Bang et al. (2007)
	Exogenous antagonists
	Tranilast	Hisanaga et al. (2009)
TRPV3	Endogenous agonist:
	Farnesyl pyrophosphate	Bang et al. (2010a)
	Endogenous antagonists:
	Isopentenyl pyrophosphate	Bang et al. (2011)
	Resolvin D1	Bang et al. (2010b)
	Exogenous agonists
	Camphor	Moqrich et al. 2005)
	Menthol	Macpherson et al. 2006)
	Eugenol, thymol, carvacrol	Xu et al. (2006a)
	6-t-butyl-m-cresol, dihydrocarveol, (+)-borneol	Vogt-Eisele et al. (2007)
	2-APB	Hu et al. (2004)
	Incensole acetate	Moussaieff et al. (2008)
	Exogenous antagonist:
	GRC15300	Khairatkar-Joshi et al. (2010)
TRPV4	Endogenous agonists:
	Citric acid	Suzuki et al. (2003)
	5,6-and 8,9-epoxyeicosatrienoic acid	Watanabe et al. (2003)
	Dimethylallyl pyrophosphate	Bang et al. (2012)
	Endogenous antagonist:
	Resolvin D1	Bang et al. (2010b)
	Exogenous agonists:
	4α-Phorbol 12, 13-dedecanoate	Klausen et al. (2009)
	Bisandrographolide	Smith et al. (2006)
	Apigenin	Ma et al. (2012a)
	GSK1016790A*2009	Thorneloe et al. (2008)
	RN-1747	Vincent et al. (2009)
	Exogenous antagonists:
	HC-067047*2009	Everaerts et al. (2010a)
	RN-1734	Vincent et al. (2009)
	GSK2193874	Huh et al. (2012; Thorneloe et al. (2012)
TRPV6	Exogenous antagonist:
	2-APB	Kovacs et al. (2012)
TRPC3	Exogenous antagonists:
	Pyr3*2009	Kiyonaka et al. (2009)
	Pyr10	Schleifer et al. (2012)
TRPC4	Exogenous antagonist:
	ML204	Miller et al. (2011)
TRPC5	Endogenous agonist:
	Lysophosphatidylcholine	Flemming et al. (2006)
	Sphingosine-1-phosphate	Xu et al. (2006b)
	Exogenous agonist:
	Rosiglitazone	Majeed et al. (2011a)
	Progesterone and neurosteroids	Majeed et al. (2011b)
TRPC6	Endogenous agonist:
	20-Hydroxyeicosatetraenoic acid	Basora et al. (2003)
	Exogenous agonists:
	Hyperforin	Leuner et al. (2007)
	2,4-Diacylphloroglucinol	Leuner et al. (2010)
	Exogenous antagonist:
	GsMTx-4	Spassova et al. (2006)
TRPM2	Endogenous agonists:
	ADP-ribose	Perraud et al. (2005)
	Cyclic ADP-ribose	Kolisek et al. (2005)
	Exogenous antagonists:
	N-(p-amylcinnamoyl)anthranilic acid	Kraft et al. (2006)
	Clotrimazole, econazole	Hill et al. (2004a)
	2-APB	Togashi et al. (2008)
	Flufenamic acid	Hill et al. (2004b; Naziroğlu et al. (2007)
TRPM3	Endogenous agonists:
	Pregnenolone sulphate	Wagner et al. (2008)
	D-erythro-sphingosine	Grimm et al. (2005)
	Endogenous antagonist:
	Progesterone	Majeed et al. (2012)
	Exogenous antagonist:
	Rosiglitazone	Majeed et al. (2011a)
	Mefenamic acid	Klose et al. (2011)
	Naringenin, hesperetin, ononetin, eriodictyol	Straub et al. (2013)
	TM3E3 (polyclonal antibody)	Naylor et al. (2008)
TRPM4	Exogenous agonist:
	BTP2	Takezawa et al. (2006)
	Exogenous antagonist:
	9-Phenanthrol	Grand et al. (2008)
TRPM5	Exogenous antagonist:
	Triphenylphosphine oxide	Palmer et al. (2010)
TRPM6	Exogenous antagonist:
	2-APB	Li et al. (2006)
TRPM7	Endogenous antagonist:
	Sphingosine	Qin et al. (2013)
	Exogenous antagonists:
	2-APB*2009	Li et al. (2006)
	Carvacrol	Parnas et al. (2009)
	Nafamostat mesilate (dependent on extracellular divalent ions)	Chen et al. (2010)
	NDGA, AA861, MK886	Chen et al. (2010)
	Waixenicin A	Zierler et al. (2011)
	FTY720	Qin et al. (2013)
	Quinine, CyPPA, dequalinium, NS8593, SKA31, UCL 1684	Chubanov et al. (2012)
TRPM8	Exogenous agonists:
	Menthol	Peier et al. (2002a)
	Linalool, geraniol, hydroxycitronellal, WS-3, WS-23, FrescolatMGA, FrescolatML, PMD38, CoolactP, Cooling Agent 10	Behrendt et al. (2004)
	Cis-and trans-p-menthane3	Bandell et al. (2004)
	CPS-368	Sherkheli et al. (2010)
	Exogenous antagonists:
	AMTB	Lashinger et al. (2008)
	BCTC	Behrendt et al. (2004)
	Benzimidazoles	Parks et al. (2011); Calvo et al. (2012)
	5-Benzyloxytryptamine	DeFalco et al. (2010)
	Compound 9l	Matthews et al. (2012)
	Tetrahydroisoquinoline 87	Tamayo et al. (2012)
	Arylglycine derivatives	Zhu et al. (2013)
TRPA1	Endogenous agonists:
	15-Deoxy-Δ12,14-PGJ2,	Materazzi et al. (2008;)
	8-Iso-PGA2, PGA2, Δ12-PGJ2	Taylor-Clark et al. (2008a)
	4-Hydroxynonenal	Trevisani et al. (2007)
	4-Oxononenal	Taylor-Clark et al. (2008b)
	Methylglyoxal	Ohkawara et al. (2012)
	Endogenous antagonists:
	Resolvin D1 Resolvin D2	Bang et al. (2010b); Park et al. (2011a)
	Exogenous agonists:
	Cinnamaldehyde, methyl salicylate, eugenol, gingerol	Bandell et al. (2004)
	Allicin, diallyl disulfide	Bautista et al. (2005)
	Δ9-tetrahydrocannabinol, Isothiocyanates	Jordt et al. (2004)
	Acrolein	Bautista et al. (2006)
	Carvacrol	Xu et al. (2006a)
	Formalin	McNamara et al. (2007)
	α,β-Unsaturated aldehydes	Andrè et al. (2008)
	Auranofin	Hatano et al. (2013)
	Capsiate	Shintaku et al. (2012)
	Curcumin	Leamy et al. (2011)
	PF-4840154	Ryckmans et al. (2011)
	Apomorphine (agonist in low micromolar range and antagonist in higher concentration)	Schulze et al. (2013)
	Cannabichromene, cannabidiol, cannabinol*2009	De Petrocellis et al. (2008; 2011)
	Exogenous antagonists:
	Camphor	Xu et al. (2005)
	Menthol	Macpherson et al. (2006)
	Thymol	Lee et al. (2008)
	HC-030031	McNamara et al. (2007)
	Chembridge-5861528	Wei et al. (2009)
	AP18	Petrus et al. (2007)
	A-967079	McGaraughty et al. (2010)
	AZ465	Nyman et al. (2013)
	GRC17536	Kaneko and Szallasi (2013)
TRPML1	Exogenous agonists:
	SF-51, ML-SA1	Shen et al. (2012)
TRPML2	Exogenous agonists:
	SID24801657, SID24787221	Saldanha et al. (2010–2009)
TRPML3	Exogenous agonists:
	SID24801657, SID24787221	Saldanha et al. (2010–2009)

^*Denotes structures shown in Figures 3 –6.

Figure 3.

Selected TRPV1 agonists: capsaicin (the pungent ingredient in hot chili peppers), resiniferatoxin (isolated from the cactus-like perennial E. resinifera), and the endocannabinoids anandamide and palmitoylethanolamide.

Few generalizations can be made about TRP channels. Some show a highly restricted tissue expression pattern (TRPA1 and TRPV1 are predominantly expressed in sensory neurons; Patapoutian et al., 2009) whereas others (TRPCs) are rather ubiquitously expressed (Singh et al., 2012). Because members of the TRP family of channels (Figure 1) share much less homology with one another compared with other ion channel families (Wu et al., 2010), the identification of highly subtype-selective compounds is likely to be more attainable.

Despite the striking progress in our understanding of TRP channel functions (see Owsianik et al., 2006; Wu et al., 2010; Li et al., 2011a), some inherent problems persist. According to Bernd Nilius, TRP channels have a fair and an ugly face (Nilius, 2013). Activation or inhibition of a TRP channel may be beneficial in one organ and, at the same time, may induce unacceptable adverse effects in another. Indeed, the clinical development of first-generation TRPV1 antagonists was halted because they caused hyperthermia and put patients at risk for scalding injuries by elevating the heat pain threshold (see Moran et al., 2011; Brederson et al., 2013). Drug discovery companies that find a way for exploiting the ‘fair face’ of TRP channels without revealing the ‘ugly face’ will be able to create a new generation of targeted therapies.

TRP channels: a brief overview

TRP channels were initially discovered in a blind strain of Drosophila (Montell and Rubin, 1989). When exposed to prolonged intense light, these spontaneously mutant fruit flies showed transient calcium influx into their photoreceptor cells; this is why the mutant gene was termed trp, ‘transient receptor potential’. This seminal finding paved the way to the discovery of the first mammalian TRP channels, called ‘canonical’ (TRPC) due to their homology to the Drosophila channel (Wes et al., 1995; Zhu et al., 1995).

Mammalian TRP channels comprise 28 members and are divided into six subfamilies: TRPC (Canonical), TRPV (Vanilloid), TRPM (Melastatin), TRPP (Polycystin), TRPML (Mucolipin) and TRPA (Ankyrin) based on their homology of amino acid sequences (Figure 1; Clapham et al., 2001; Wu et al., 2010; Nelson et al., 2011). The mucolipin and polycystin subfamilies were named after the diseases they are associated with, mucolipidosis and autosomal dominant polycystic kidney disease (ADPKD) respectively. The vanilloid subfamily was named after its founding member, the vanilloid (capsaicin) receptor TRPV1. The first melastatin channel (TRPM1) was discovered as a protein present in benign nevi and absent in malignant melanoma (Duncan et al., 1998). As of today, the ankyrin subfamily has only one member, TRPA1, which (as the name implies) is rich in ankyrin repeats at its N-terminus.

As a general rule, TRP channels have six transmembrane spanning domains (S1–S6) with a pore-forming loop between S5 and S6 (Figure 1; Wu et al., 2010). Both –NH2 and –COOH termini are located intracellularly. Many TRP channels are non-selective Ca²⁺-permeable channels with permeability ratios P_Ca/P_Na < 10. TRPM4 and TRPM5, in particular, are only permeable to monovalent cations and they do not conduct Ca²⁺ and Mg²⁺, while TRPV5 and TRPV6 are highly Ca²⁺ selective with P_Ca/P_Na > 100 (Owsianik et al., 2006). Most TRPs form functional channels as homotetramers, but heteromultimerization is frequently observed (Cheng et al., 2010). This creates a potential problem for drug discovery efforts as heteromultimers (that are not easily recreated in heterologous expression systems) may have distinct pharmacological properties.

TRP channels are ‘cellular sensors’ (Clapham, 2003) that respond to changes in the cellular environment, including temperature, stretch/pressure, chemicals, oxidation/reduction, osmolarity and pH, both acidic and alkaline (Moran et al., 2011; Nieto-Posadas et al., 2011a). Of note, a number of TRP channels are also activated by natural products, including herbs, spices, venoms and toxins (Vriens et al., 2008). For example, TRPV1 is a shared target for capsaicin (Caterina et al., 1997), jelly fish venoms and spider (tarantula) toxins (Cromer and McIntyre, 2008). A list of ligands for TRP channels is provided in Table 1. Representative chemical structures of TRP channel agonists (GSK101679A for TRPV4) and antagonists (GRC6211, AMG517, SB-705498, AZD1386, JTS-653 and ABT-102 for TRPV1; HC-067047 for TRPV4; Pyr3 for TRPC3; and AMTB for TRPM8) are shown in Figures 4 and 5. Despite decades of intensive search, only a few endogenous ligands of TRP channels have been identified, such as the endocannabinoids anandamide and palmitoylethanolamide, (structures shown in Figure 3) for TRPV1 (Zygmunt et al., 1999; Ambrosino et al., 2013)]. How TRP channels are modulated in vivo is still unknown.

Figure 4.

Small molecule TRPV1 antagonists: selected structures.

Figure 5.

Representative examples of TRPV4 agonists (GSK1016790A), TRPV4 antagonists (HC-067047), TRPV3 agonists (2-ABT), and TRPC3 inhibitors (Pyr3).

Some TRPM channels like TRPM2 are unique in that they contain a functional nucleoside diphosphate linked to some other moiety/ADP ribose domain, as well as a kinase domain that bears some resemblance to PKA (see Eisfeld and Lückhoff, 2007). In other words, these TRPMs combine features of ion channels and enzymes and are thus referred to by some as ‘chanzymes’ (Montell, 2003). TRPM2 functions as a cellular redox (oxidative stress) sensor and has been implicated in the pathogenesis of bipolar disorder, diabetes, as well as cardiovascular and neurodegenerative disorders (Jiang et al., 2010). Indeed, a mutant TRPM2 (Pro1018Leu) has been linked to the Guamanian amyotrophic lateral sclerosis (ALS-G)/Parkinsonism-dementia complex (Hermosura et al., 2008).

Importantly, TRP channels are also stimulated by intracellular Ca²⁺ increase induced by the activation of GPCRs and mediate downstream signalling. Furthermore, the activity of TRP channels is modulated by various intracellular molecules including phosphatidylinositol 4,5-bisphosphate (PIP₂), DAG, ATP and calmodulin (see Wu et al., 2010). Indeed, TRPCs are subdivided into two groups, TRPC1, C4 and C5 and TRPC3, C6, C7, depending on DAG-sensitivity (Wu et al., 2010). For some TRP channels, like TRPV1, phosphorylation by PKs and dephosphorylation by phosphatases provide important contribution to setting the channel activity (reviewed in Szallasi et al., 2007). Indeed, phosphorylation of TRPV1 by PKC is now thought to represent a crucial biochemical mechanism by which TRPV1 is sensitized during inflammation to cause thermal hyperalgesia (Jeske et al., 2009). Moreover, TRP channels were shown to interact with a growing number of intracellular proteins to form ‘signalplexes’ and ‘channelosomes’ (Planells-Cases and Ferrer-Montiel, 2007). Such interactions are now believed to be important for TRP channel trafficking, positioning and activity (Goswami, 2012). Recent research suggests that manipulation of the interaction between TRP channels and their regulatory proteins can be exploited for therapeutic purposes. A promising example of this approach, to prevent TRPV1 sensitization by blocking the interplay between TRPV1 and the scaffolding protein AKAP79 (Btesh et al., 2013; Fischer et al., 2013), will be discussed later under ‘pain and TRP channels’.

Activation of TRP channels allows cations pass through the membrane and depolarize cells, leading to a wide range of cellular responses. Stimulated by a broad range of stimuli and expressed probably in all the cells in the body (Nilius, 2013), TRP channels are thought to play diverse physiological roles. Besides, extensive research in the field has demonstrated that TRP channels are involved in a number of diseases affecting the peripheral and CNS (Vennekens et al., 2012; Morelli et al., 2013), the respiratory (Preti et al., 2012), genito-urinary (Skryma et al., 2011), gastrointestinal (GI; Holzer, 2011), cardiovascular (Watanabe et al., 2013) and immune systems (Schwartz et al., 2007; Smith and Nilius, 2013), as well as in metabolic disorders including obesity and diabetes (Suri and Szallasi, 2008; Zhu et al., 2011).

The direct link between TRP channels and human diseases have been revealed by human genetic studies demonstrating that mutations in TRP genes are causally associated with hereditary diseases, the so-called ‘TRP channelopathies’ (Nilius and Owsianik, 2010). Representative examples of these channelopathies include focal segmental glomerular sclerosis (FSGS), ADPKD and scapuloperoneal spinal muscular atrophy, which are linked to TRPC6 (Winn et al., 2005), TRPP2 (Igarashi and Somlo, 2002) and TRPV4 (Auer-Grumbach et al., 2010) respectively. Knockout and transgenic animal studies also revealed a pathogenic role for both the absence and hyperactivity of TRP channels. For example, TRPC3 (−/−) mice show defects in motor coordination and walking behaviour (Hartmann et al., 2008) whereas transgenic mice overexpressing TRPC6 in their heart develop massive cardiac hypertrophy (Kuwahara et al., 2006).

In summary, there is strong experimental and clinical evidence to substantiate TRP channels as appealing drug targets and a number of molecules targeting TRP channels have already advanced to clinical trials (Moran et al., 2011; Brederson et al., 2013; Table 2). Later, we provide an overview of TRP channels and diseases and discuss potential approaches for therapeutic intervention.

Table 2.

Drugs targeting TRP channels in clinical development

Action	Drug	Company	Therapy Area	Highest development status	ClinicalTrials.gov identifier
TRPV1 agonist	capsaicin	Not Assigned	Pain	Launched
TRPV1 agonist	NGX-4010	Acorda Therapeutics Inc/Astellas Pharma Inc	Postherpetic neuralgia	Launched
TRPV1 agonist	zucapsaicin	Sanofi-Aventis Canada Inc	Osteoarthritis	Registered
TRPV1 agonist	zucapsaicin	Winston Pharmaceuticals Inc	Cluster headache	Phase 3	NCT00033839
TRPV1 agonist	MCP-101 (resiniferatoxin)	Mt Cook Pharma	Overactive bladder	Phase 2	N/A
TRPV1 antagonist	DWP-05195	Daewoong Pharmaceutical Co Ltd	Neuropathic pain	Phase 2	NCT01557010
TRPV1 antagonist	XEN-D0501	Provesica Ltd	Overactive bladder	Phase 2	N/A
TRPV1 siRNA	SYL-1001	Sylentis Sau	Ocular pain	Phase 2	NCT01776658
TRPV1 antagonist	Mavatrep	Johnson & Johnson Pharmaceutical Research & Development LLC	Osteoarthritis/Pain	Phase 1	NCT00933582
					NCT01006304
TRPV1 antagonist	PHE-377	PharmEste SRL	Neuropathic pain	Phase 1	N/A
TRPV1 antagonist	MR-1817	Mochida Pharmaceutical Co Ltd	Pain	Phase 1	NCT00960180
TRPV1 antagonist	PAC-14028	Pacific Pharmaceuticals Co Ltd	Atopic dermatitis/IBD	Phase 1	NCT01638117
TRPV1 antagonist	SB-705498	GlaxoSmithKline plc	Pruritus	Phase 1	NCT01673529
TRPV3 antagonist	GRC-15300	Glenmark Pharmaceuticals Ltd/Sanofi	Neuropathic pain/Osteoarthritis	Phase 2	NCT01463397
TRPM8 agonist	Menthol	Not assigned	Carpal tunnel syndrome/Neck pain	N/A	NCT01716767
					NCT01542827
TRPM8 agonist	D-3263	Dendreon Corp	Cancer	Phase 1	NCT00839631
TRPA1 antagonist	GRC-17536	Glenmark Pharmaceuticals Ltd	Diabetic peripheral neuropathy/Respiratory disorders	Phase 2	NCT01726413
TRPA1 antagonist	CB-625	Cubist Pharmaceuticals/Hydra Biosciences	Inflammatory disease/ Pain	Phase 1	N/A

Pain and TRP channels

A number of TRP channels (including TRPV1, V3 and V4, TRPA1, TRPM3 and M8, TRPC1, C3 and C6) are expressed in nociceptive sensory neurons. Extensive research with genetically modified animals and pharmacological agents has confirmed that these TRP channels are involved in the generation and transduction of pain and thus represent promising targets for the development of novel analgesic agents (see Patapoutian et al., 2009; Moran et al., 2011; Brederson et al., 2013).

TRPV1

A subset of nociceptive neurons with somata in sensory (dorsal root and trigeminal) ganglia is distinguished by its unique sensitivity to capsaicin (Szallasi and Blumberg, 1999). The initial excitation by capsaicin of these neurons is followed by a lasting refractory state (traditionally referred to as desensitization) in which the cells are unresponsive not only to a repeated capsaicin challenge, but also to various unrelated stimuli like noxious heat and acids (see Szallasi and Blumberg, 1999). Thus, desensitization by capsaicin has a clear therapeutic potential. The receptor for capsaicin was identified as TRPV1 (Caterina et al., 1997) and accumulating evidence suggests a crucial role for TRPV1 in pain sensation (see Szallasi et al., 2007; Gomtsyan and Faltynek, 2010). First, TRPV1 is activated by multiple painful stimuli including noxious heat, pungent chemicals (capsaicin and jelly fish venom), and protons (Szallasi et al., 2007). In addition, TRPV1 can be activated by voltage, lipids and phosphorylation (Pingle et al., 2007). Second, TRPV1-deficient mice show reduced thermal hyperalgesia in response to inflammatory mediators such as bradykinin and/or NGF (Caterina et al., 2000; Davis et al., 2000; Chuang et al., 2001). In addition, oleoylethanolamide, an endogenous TRPV1 agonist (Ahern, 2003), induces visceral pain-related behaviour in mice that is inhibited by the TRPV1 antagonist capsazepine and is absent in TRPV1-null animals (Wang et al., 2005). Third, pharmacological blockade or knockdown of TRPV1 displays analgesic activity in various preclinical pain models, including arthritic (Szabó et al., 2005; Joshi and Honore, 2010) and cancer pain (Jimenez-Andrade and Mantyh, 2010).

Somewhat unexpectedly, genotyping studies have so far failed to identify any TRPV1 polymorphism associated with neuropathic pain, although TRPV1 variants were correlated with altered somatosensory function in patients with neuropathic pain (Binder et al., 2010). Parenthetically, the TRPV1 585 Ile-Ile genotype appears to lower the risk for developing painful knee osteoarthritis (Valdes et al., 2011). But not all TRPV1 variants are harmless. For example, in a large European study, six TRPV1 gene SNPs appeared to confer higher risk for chronic cough [Smit et al., 2012; although, in a different study, the loss-of-function TRPV1 variant I585V was associated with a lower risk for childhood asthma (Cantero-Recasens et al., 2010).

Capsaicin-containing creams (e.g. Zostrix, 0.075%) have been used for decades for the treatment of chronic painful conditions such as diabetic neuropathy (Knotkova et al., 2008). Despite their popularity, controlled clinical studies found no evidence that these creams had greater analgesic potency than placebo (see Szallasi and Sheta, 2012).

To increase the exposure of cutaneous nerve endings to capsaicin, occlusive patches (NGX-4010, Qutenza) and liquid formulations (NGX-1998, 20% capsaicin) were developed by Neuroges-X (San Mateo, CA, USA; Bley, 2012). Although in 2010, Qutenza was approved to treat post-herpetic neuralgia in the USA, the sales of this $700 pain patch never matched the company's expectations. In 2012, after the US Food and Drug Administration rejected Neuroges-X's request to extend the use of Qutenza to HIV-associated peripheral neuropathy, the company ceased operations and reached a tentative deal to sell Qutenza and the investigational liquid formulation NGX-1998 to Acorda (Ardsley, NY, USA; http://www.researchviews.com/healthcare/pharma/DealReports.aspx?sector=Pharma&DealID=191932).

Resiniferatoxin is currently undergoing clinical trials at the National Cancer Institute in patients with intractable cancer pain as a ‘molecular scalpel’ to achieve permanent analgesia (NCT00804154). In preclinical models of chronic pain, intrathecal resiniferatoxin induces a lasting analgesic effect by selectively ablating TRPV1-expressing sensory neurons in the dorsal root and trigeminal ganglia (Iadarola and Gonnella, 2013). In client-owned dogs with severe osteosarcoma pain, intrathecal resiniferatoxin was well-tolerated and effective: it provided significant pain relief and restored ambulation for several months after a single administration (Brown et al., 2005; Iadarola and Gonnella, 2013). It is hoped that intrathecal resiniferatoxin will be a good alternative to narcotic analgesics in some cancer patients with localized pain, such as pain caused by bone metastasis.

After the cloning of TRPV1, there was a great deal of enthusiasm in the pharmaceutical industry to develop small-molecule TRPV1 antagonists as analgesic agents. Indeed, a number of TRPV1 antagonists including SB-705498 (GlaxoSmithKline), AMG517 (Amgen), AZD1386 (AstraZeneca), GRC-6211 (Lilly/Glenmark), MK-2295 (Merck/Neurogen), ABT-102 (Abbott) and PHE377 (PharmEste) have been advanced to Phase I and II clinical studies for indications related to pain (representative structures are shown in Figure 4; Moran et al., 2011; Brederson et al., 2013). The enthusiasm, however, was soon tempered by unforeseen adverse effects. Some TRPV1 antagonists (AMG517) caused marked hyperthermia, prompting their withdrawal from the clinical trials, whereas others (MK-2295) blunted noxious heat perception, putting patients at risk for scalding injuries (Moran et al., 2011; Brederson et al., 2013).

The magnitude of hyperthermia seems to vary depending on the chemical structure. While AZD1386 modestly increased body temperature (∼0.4 °C on average) in patients with gastroesophageal reflux disease (GERD; Krarup et al., 2011), AMG517 caused a lasting (1–4 days) and marked hyperthermia response (up to 40.2°C) in human volunteers (compare AZD1386 and AMG517 structures in Figure 4; Gavva et al., 2008). In preclinical studies (in rodents and dogs), PHE377 (structure undisclosed) was devoid of any effect on body temperature at doses at which it inhibited both thermal and mechanical hyperalgesia (http://www.pharmeste.com/repository/contenuti/paragrafi/file/PharmEste_Leaflet_2012.PDF).

The site that mediates the hyperthermic action of TRPV1 antagonists is still hotly debated. In rodents, capsaicin evokes transient hypothermia (presumably by activating cooling mechanisms after tricking the animals into believing that they are hot), followed by a loss of the animals' ability to regulate their body temperature (rats desensitized to capsaicin develop hyperthermia when placed in a hot chamber; Szallasi and Blumberg, 1999). The effects of capsaicin on thermoregulation were linked to the CNS (Hajós et al., 1985). However, TRPV1 antagonists that do or do not enter the CNS are comparable in their ability to elevate body temperature, making a CNS target extremely unlikely (Cui et al., 2006). Consequently, it was postulated that TRPV1 in the periphery has an endogenous tone that is essential for maintaining normal body temperature (Gavva, 2008). However, rodents whose TRPV1 has been eliminated by genetic recombination (TRPV1 kncok out mice) or chemical ablation (neonatal capsaicin treatment) do not develop hyperthermia.

Similar to the hyperthermic response, the magnitude of blunted heat perception also seems to depend on the TRPV1 antagonist pharmacophore. The increase in heat pain threshold was first noted after the administration of 400 mg of SB-705498 to healthy human volunteers (Chizh et al., 2007). Importantly, unlike the febrile reaction that disappeared upon repeated dosing, the impaired thermal sensitivity persisted during the whole course of the study (Chizh et al., 2007). Impaired heat perception was also observed with MK-2295 (Eid, 2011), ABT-102 (Rowbotham et al., 2011) and AZD1386 (Krarup et al., 2011). Some volunteers receiving MK-2295 perceived potentially harmful temperature as innocuous (Eid, 2011). Indeed, minor (1st and 2nd degree) burns were reported in some clinical study subjects. Interestingly, another clinical study (XEN-D0501) found no evidence of scalding injuries in the study participants (Round et al., 2011).

The heat sensor in TRPV1 is clearly distinct from the capsaicin and proton recognition sites (Szolcsányi and Sándor, 2012) and a new generation of modality-selective TRPV1 antagonists that cause neither hyperthermia nor impaired pain heat perception have been proposed (Szolcsányi and Sándor, 2012; Brederson et al., 2013).

An attractive alternative approach to circumvent the side effects of TRPV1 antagonists is to target TRPV1 in diseased, but not in healthy, tissues (Szallasi and Blumberg, 2006). Phosphorylation of the TRPV1 protein by PKA and PKC is believed to play a crucial role in inflammatory sensitization (reviewed in Szallasi et al., 2007). The interaction between both kinases and TRPV1 depends on the scaffolding protein AKAP79 (Zhang et al., 2008). Recently, specific residues in TRPV1 and AKAP79 were discovered by site-directed mutagenesis experiments where these proteins interact (Btesh et al., 2013; Fischer et al., 2013). This information allowed the design and synthesis of peptides that can block the interaction between TRPV1 and AKAP79. In mice, these blocking peptides prevented the development of inflammatory thermal hyperalgesia but it remains to be seen if a similar strategy can be successful in chronic pain patients.

Neither AMG-9810 (a TRPV1 antagonist) nor HC-030031 (a TRPA1 blocker, see later) relieved ongoing pain in a mouse model of osteoarthritis, although both ameliorated thermal hyperalgesia (Okun et al., 2012). Consisitent with these findings, in a randomized, double-blinded, prospective clinical trial with client-owned dogs suffering from severe hip osteoarthritic pain, the TRPV1 antagonist ABT-116 showed only marginal analgesic activity over placebo (Malek et al., 2012). ABT-116 did not attenuate lameness in dogs with experimentally induced urate synovitis either at doses at which it caused seriously high rectal temperatures (Cathcart et al., 2012). The striking difference in the analgesic activity of TRPV1 antagonists between rodent (where it potently reduces experimental osteoarthritic pain; Honore et al., 2005; Puttfarcken et al., 2010) and canine (without clinical benefits, see earlier) models of human osteoarthritic pain is puzzling and concerning. Indeed, osteoarthritic pain is a major clinical indication for TRPV1 antagonists and a number of TRPV1 antagonists have entered clinical trials for this indication (Table 2), but as yet, the results have not been disclosed.

TRPA1

Unlike TRPV1 (which shows distinct structure-activity relations in its ligand-binding properties, hence the original name ‘vanilloid receptor’), TRPA1 is activated by a wide range of irritant natural products, including allyl isothiocyanate (Jordt et al., 2004; Capasso et al., 2012), cinnamaldehyde (Bandell et al., 2004) and allicin (Bautista et al., 2005) found in mustard oil, cinnamon and garlic respectively. TRPA1 is also targeted by environmental irritants found in tear gas, exhaust fumes, household cleaning agents and cigarette smoke; examples include acrolein (Bautista et al., 2006), formalin (McNamara et al., 2007) and α,β-unsaturated aldehydes (Andrè et al., 2008). Of note, TRPA1 binds umbellulone, an active ingredient in the Californian ‘headache tree’ Umbellularia californica (Nassini et al., 2012a). Indeed, activation of TRPA1 expressed on meningeal afferents was implicated in the pathomechanism of migraine (Edelmayer et al., 2012). Somewhat surprisingly for an ‘irritant receptor,’ TRPA1 is also activated by the non-pungent capsaicin analogue, capsiate (Shintaku et al., 2012), as well as the non-psychotropic cannabinoid, cannabichromene (see Figure 6 for structure; De Petrocellis et al., 2008;2011). This is interesting because cannabichromene is thought to play a pivotal role in the anti-inflammatory and analgesic activity of medical marijuana and cannabichromene did ameliorate experimental murine colitis (Romano et al., 2013). TRPA1 also acts as a receptor for reactive oxygen species (ROS; Bessac et al., 2008), but it is debated if TRPA1 can be activated by noxious cold. In general, reactive chemicals activate TRPA1 by inducing covalent modification of cysteines in the N-terminus (Nilius et al., 2011).

Figure 6.

Selected plant cannabinoids that target TRP channels: cannabichromene (TRPA1 agonist) and cannabidiol (TRPM8 antagonist).

TRPA1 is well established as a pain sensor. A gain-of-function point mutation in TRPA1 (N855S) was identified as the cause of familial episodic pain syndrome, a rare human pain disorder characterized by severe upper body pain triggered by fasting and physical stress (Kremeyer et al., 2010). TRPA1 is primarily expressed in sensory neurons (where it is highly coexpressed with TRPV1), but growing evidence suggests that functional TRPA1 is also present in non-neuronal tissues such as heart, small intestine, lung and pancreas (Stokes et al., 2006). As discussed later under respiratory disorders, TRPA1 expressed by lung fibroblasts (Mukhopadhyay et al., 2011) might play a role in the pathogenesis of asthma and COPD (Nassini et al., 2012b).

TRPA1 expression in sensory neurons appears to be plastic and neuropathic injury increased neuronal expression of TRPA1 in humans (Anand et al., 2008). Cyclopentenone PGs, metabolites of PGs enhancing pain sensation, cause a robust calcium response in dorsal root ganglion neurons and induce pain behaviour in wild-type mice, but not in TRPA1-deficient mice (Materazzi et al., 2008). Interestingly, the pain phenotype of TRPA1 knockout and knock-down animals are different: inflammation-induced mechanical hyperalgesia is reduced in the knock-down mice, but not in the knockouts (see Garrison and Stucky, 2011; Nilius et al., 2011). This implies the existence of a compensatory mechanism that takes over the function of the missing TRPA1 in the knockout animals and restores mechanical hyperalgesia during inflammation.

Pharmacological inhibition of TRPA1 with HC-030031, a highly selective TRPA1 antagonist, attenuated formalin-induced pain (McNamara et al., 2007) and reversed mechanical hypersensitivity following complete Freund's adjuvant (CFA) treatment. HC-030031 also displayed analgesic activity in the spinal nerve ligation model of neuropathic pain (Eid et al., 2008). CHEM-5861528, a derivative of HC-030031, alleviated mechanical hyperalgesia in a rat model of diabetic neuropathic pain (Wei et al., 2009; Koivisto et al., 2012). Of note, methylglyoxal (an endogenous carbonyl compound that is produced in large amounts during hyperglycemic conditions) activates human TRPA1 (Ohkawara et al., 2012). Furthermore, TRPA1 has been implicated in migraine (Edelmayer et al., 2012), dental pain (Haas et al., 2011), chemotherapy-induced neuropathic pain (Nassini et al., 2013) and colicky pain of GI origin (Blackshaw et al., 2013).

The Abbott TRPA1 antagonist A-967079 attenuated both evoked and spontaneous firing recorded from wide dynamic range (WDR) spinal cord neurons during CFA-induced inflammation (McGaraughty et al., 2010) without having any effect on body temperature (Chen et al., 2011). However, in rats with osteoarthritic pain, A-967079 had no effect on spontaneous WDR firing (though it blocked evoked mechanical hyperalgesia), suggesting (somewhat disappointingly) that TRPA1 blockade may not alleviate the on-going, spontaneous ‘nagging’ pain in patients with osteoarthritis (McGaraughty et al., 2010).

There is good evidence linking TRPA1 to the cold allodynia that develops during ciguatera (Vetter et al., 2012) or following chemotherapy (Nassini et al., 2013). In some studies, TRPA1 was directly activated by noxious cold (Story et al., 2003). Paradoxically, TRPA1 expressed on polymodal C-fibres appears to be activated by hot temperatures (Hoffmann et al., 2013). It is tempting to speculate that this mechanism is responsible for the development of inflammatory thermal hyperalgesia which is absent in the TRPA1 knockout mice (P. Reeh, pers. comm.).

To date, two TRPA1 antagonists have reached clinical stage of development, GRC1753 (Glenmark) for chronic pain and CB-625 for acute surgical pain (Cubist Pharmaceuticals Inc., Lexington, MA, USA; Hydra Biosciences, Inc., Cambridge, MA, USA).

TRPV3

TRPV3 is abundantly expressed in keratinocytes where it is thought to serve various functions (Nilius and Bíró, 2013). Keratinocytes release IL-1, a pro-inflammatory cytokine, in response to eugenol, a non-selective TRPV3 agonist (Xu et al., 2006a). TRPV3 expression is significantly increased in keratinocytes in patients with breast pain (Gopinath et al., 2005); by contrast, TRPV3 is decreased in keratinocytes in patients with diabetic neuropathy (Facer et al., 2007). In addition, TRPV3 was significantly increased in brachial plexus nerves collected from patients with traumatic nerve injury (Facer et al., 2007). Interestingly, Olmsted syndrome patients (a genetic disorder caused by a gain-of-function TRPV3 mutation; Lin et al., 2012) suffer from intense itching but not pain.

TRPV3 is activated by warm temperatures in the range of 31–39°C and its activity is enhanced during repetitive heat stimulations (Xu et al., 2002; Peier et al., 2002b). TRPV3-null mice showed marked deficits in responses to innocuous and noxious heat (Moqrich et al., 2005). GRC15300, a potent, selective, orally available TRPV3 antagonist demonstrated efficacy in inflammatory and neuropathic pain models, and this compound is being investigated in clinical trials (Khairatkar-Joshi et al., 2010).

TRPM8, TRPV4, TRPM3 and TRPCs

TRPM8 is expressed in nociceptive Aδ and C fibres that are cold sensitive (McKemy et al., 2002; Kobayashi et al., 2005). TRPM8 is activated by cold temperatures in the range of 8–28°C, as well as by cooling compounds such as menthol and icilin (McKemy et al., 2002; Peier et al., 2002a). Topical menthol has been tried clinically as an analgesic in patients with carpal tunnel syndrome (NCT01716767) and neck pain (NCT01542827). Naturally occurring TRPM8 antagonists include the plant cannabinoids cannabidiol (see Figure 6 for structure), cannabinol and cannabiogerol (De Petrocellis et al., 2008;2011). Studies with TRPM8-null mice showed decreased sensitivity to cold temperature, as well as attenuated hypersensitivity to cold after nerve injury or inflammation (Bautista et al., 2007; Colburn et al., 2007). Synthetic TRPM8 antagonists were analgesic in a chronic constriction injury-induced model of neuropathic pain in rats (Parks et al., 2011; Calvo et al., 2012) but, as yet, no TRPM8 antagonist has advanced to clinical trials.

TRPV4 as a pain target is highly controversial. Intraplantar injection of the endogenous TRPV4 activator dimethylallyl pyrophosphate elicits nociceptive flinches (Bang et al., 2012). Furthermore, TRPV4-deficient mice demonstrated decreased pain behaviour in inflammatory pain models (Todaka et al., 2004; Alessandri-Haber et al., 2006), as well as models of painful peripheral neuropathy (Alessandri-Haber et al., 2008). Basal visceral nociception and TRPV4 agonist-induced visceral hypersensitivity were reduced by intervertebral injection of TRPV4-targeted siRNA (Cenac et al., 2008). That said, none of the numerous gain-of-function TRPV4 channelopathies has a painful phenotype.

TRPM3-deficient mice showed impaired behavioural response to noxious heat and failed to develop inflammatory heat hyperalgesia (Vriens et al., 2011). Naturally occurring TRPM3 blockers include the citrus fruit flavanones, naringenin and hesperetin (Straub et al., 2013). Of TRPC channels expressed by sensory neurons, TRPC5 appears to be the most interesting given its postulated role in cold-sensation (Zimmermann et al., 2011).

Respiratory disorders and TRP channels

The mammalian respiratory tract is densely innervated by sensory afferent fibres whose activation by irritant and/or inflammatory stimuli evokes a myriad of central and peripheral protective reflex responses, including cough, mucus secretion and bronchospasm (Canning, 2006). The pulmonary chemoreflex is a triad of bradycardia, bradypnea and hypotension. Of TRP channels expressed in these afferents, TRPA1 and TRPV1 have attracted the most attention as sensors of environmental irritants and reactive chemicals that threaten airway function and integrity.

TRPA1

Hypochlorite (the oxidizing mediator of chlorine) and hydrogen peroxide (a ROS) activate TRPA1 in chemosensory neurons in mice to cause respiratory depression, as well as pain behaviour, both of which were attenuated in TRPA1-deficient mice (Bessac et al., 2008). Ozone, one of the major air pollutants, stimulated a subset of nociceptive sensory neurons isolated from vagal ganglia of wild-type mice, but not those from TRPA1 knockout mice (Taylor-Clark and Undem, 2010).

TRPA1 is also targeted by endogenously generated pro-inflammatory ligands as exemplified by nitro-oleic acid, a nitrated phospholipid produced during inflammation (Taylor-Clark et al., 2009). Cigarette smoke is the major cause of COPD and is among the most prevalent triggers of asthma. Crotonaldehyde and acrolein, two main components of cigarette smoke, evoked Ca²⁺ influx in cultured guinea pig jugular ganglia neurons and promoted contraction of isolated guinea pig bronchi (Andrè et al., 2008). These responses were abolished by the selective TRPA1 antagonist HC-030031 (Andrè et al., 2008). Moreover, inhalation of acrolein and other irritant TRPA1 agonists causes cough both in guinea pigs and human volunteers (Andrè et al., 2009; Birrell et al., 2009). Genetic deletion and/or pharmacological blockade of TRPA1 inhibited leukocyte infiltration in the airways, reduced cytokine and mucus production, and almost completely abolished airway hyperreactivity to contractile stimuli in a murine ovalbumin model of asthma (Caceres et al., 2009).

In the lung, TRPA1 is also expressed by fibroblasts (Mukhopadhyay et al., 2011). Most recently, this non-neuronal TRPA1 has been linked to non-neurogenic airway inflammation (Nassini et al., 2012b). This is important because neurogenic inflammation, although predominant in preclinical models of asthma, plays lesser, if any, role in the human disease. Indeed, tachykinin receptor (NK1, 2 and 3) antagonists were without any clear benefit in clinical trials for asthma. Taken together, these findings suggest that selective TRPA1 antagonist may provide therapeutic benefits in respiratory diseases characterized by airway inflammation, such as asthma and COPD (Belvisi et al., 2011; Preti et al., 2012).

TRPV1

TRPV1 is another key player in the control of airway sensitivity. Indeed, inhaled capsaicin evokes multiple protective reflex responses in humans, including cough, sneezing and fluid secretion (Szallasi and Blumberg, 1999). In rodents, capsaicin also evokes the pulmonary chemoreflex, which is believed to represent the major dose-limiting factor for acute capsaicin administration (in desensitization studies, capsaicin needs to be given in increasing doses in consecutive days to avoid potentially fatal respiratory depression; Szallasi and Blumberg, 1999). The capsaicin inhalation test is broadly used to identify a subset of chronic cough patients with airway sensory hyperreactivity (Ternesten-Hasséus et al., 2008). It is believed that these patients have overactive sensory nerves responsible for the airway symptoms and may benefit from inhaled TRPV1 (and/or TRPA1) antagonists (Millqvist, 2011). It is worth mentioning here that fatal asthma attacks were reported in asthma patients following incidental capsaicin inhalation (see Szallasi and Blumberg, 1999).

Hydrogen sulfide evokes neuropeptide release from isolated guinea pig airway tissue; it also contracts the guinea pig bronchus (Trevisani et al., 2005). These effects are reduced by capsaicin desensitization or by the TRPV1 antagonist, capsazepine (Trevisani et al., 2005). In anaesthetized guinea pigs, intratracheal instillation of hydrogen sulfide increases the total lung resistance and evokes neurogenic inflammation (airway plasma protein extravasation): these effects are reduced by capsazepine (Trevisani et al., 2005). TRPV1 activators induce action potential discharge in murine vagal C-fibre terminals and this response was absent in TRPV1-deficient C-fibres (Kollarik and Undem, 2004). Pharmacological inhibition of TRPV1 significantly inhibited airway hyperresponsiveness to histamine in non-anaesthetized, ovalbumin-sensitized guinea pigs (Delescluse et al., 2012). In addition, a recent study demonstrated that PGE₂ and bradykinin, two well-described endogenous inflammatory mediators, activated isolated guinea pig sensory ganglia and evoked cough in guinea pigs. Interestingly, effective blockade of this cough response required the simultaneous antagonism of both TRPV1 and TRPA1 receptors (Grace et al., 2012).

Several lines of evidence implicate TRPV1 in the pathomechanism of chronic cough (see Spina and Page, 2013). The concentration at which inhaled capsaicin evokes coughing is markedly reduced in a subpopulation of chronic cough patients. These patients show increased TRPV1 expression in their airway nerves (Groneberg et al., 2004). Furthermore, TRPV1 gene polymorphism was associated with cough sensitivity among subjects without asthma (Smit et al., 2012). Nevertheless, TRPV1 as a target for anti-tussive drugs remains controversial because inhaled SB-705498, a selective TRPV1 antagonist, failed to ameliorate spontaneous coughing in chronic cough patients (C. Page, pers. comm.).

TRPM8

TRPM8, a cold-sensing TRP channel, is expressed in a subset of autonomic afferent nerves innervating the bronchopulmonary system and activation of these nerves may increase airway resistance (Xing et al., 2008). If so, TRPM8 activation may be associated with cold-induced exacerbation of asthma and other pulmonary disorders (Xing et al., 2008). On the other hand, respiratory irritant responses evoked by vapours containing cigarette smoke constituents (acrolein, acetic acid or cyclohexanone) were alleviated by menthol, and the effect of menthol was reversed by the TRPM8 antagonist AMTB (see Figure 5 for structure; Willis et al., 2011). Indeed, menthol is added to some cigarette brands to minimize airway irritation. Furthermore, nasal application of TRPM8 agonists significantly increased the threshold of capsaicin-induced cough responses in human volunteers (Buday et al., 2012). Clearly, TRPM8 activation can be both beneficial (for example, it may reduce airway irritancy) and harmful (it may exacerbate asthma), depending on the patient.

The role of TRPM8 in airways extends beyond the sensory nerves. In bronchial epithelium of patients with COPD, TRPM8 expression is markedly increased and stimulation with cold or menthol causes MUC5AC expression. MUC5AC expression is reduced by TRPM8 shRNA in normal human bronchial epithelial cells (Li et al., 2011a). Considering that cold is one of the key triggers of COPD exacerbation and the enhanced mucus secretion contributes to morbidity of COPD by plugging airways and causing recurrent infection, TRPM8 may play an important role in the development of COPD and maybe also asthma.

TRPV4

The link between TRPV4 and human pulmonary disease was initially established by the discovery of TRPV4 gene polymorphism in COPD patients (Zhu et al., 2009). Interestingly, one of these COPD-predisposing TRPV4 variants has a gain-of-function phenotype with enhanced Ca²⁺ influx and MMP-1 release evoked by diesel exhaust particles, indicating that altered activity of TRPV4 could drive the pathogenesis of COPD in some patients (Li et al., 2011b). Normal TRPV4 activity may maintain ciliary movements and cilia on bronchial epithelial cells are essential for airway clearance. The Ca²⁺ overload secondary to this gain-of-function mutation is thought to impair ciliary function, leading to accumulation of harmful airborne particles in the lungs. Indeed, impaired ciliary movement (‘ciliopathy’) is an early sign of COPD (http://weill.cornell.edu/news/releases/wcmc_2012/07_12_12b.shtml).

TRPV4 is also implicated in the pathogenesis of pulmonary oedema caused by high pulmonary venous pressure secondary to heart failure. TRPV4 activation evoked by elevated vascular pressure induces a marked increase in pulmonary endothelial cell permeability (Jian et al., 2008). This observation is in line with the finding that systemic administration of a TRPV4 agonist elicited increased pulmonary vascular permeability, vascular haemorrhage, and circulatory collapse as a result of profound disruption of the endothelial permeability barrier (Willette et al., 2008). Conversely, TRPV4 inhibition prevented the increased vascular permeability and resultant pulmonary oedema induced by elevated pulmonary venous pressure in isolated rodent and canine lungs (Thorneloe et al., 2012). Of note, the expression of TRPV4 in the pulmonary vasculature is enhanced in lung sections obtained from heart failure patients (Thorneloe et al., 2012). Collectively, these observations imply a therapeutic benefit for TRPV4 blockade in heart failure patients with pulmonary oedema.

TRPC6

Lung ischaemia–reperfusion is another cause of pulmonary oedema. The involvement of TRPC6 in this condition is implied by the finding that TRPC6-deficient mice fail to develop oedema following lung ischaemia–reperfusion (Weissmann et al., 2012). It was suggested that TRPC6 is activated by DAG generated in a sequence of biochemical events starting from superoxide production by NADPH oxidase 2 (NOX2) and leading to elevated vascular permeability.

The role of TRPC6 in pulmonary system extends beyond oedema. For example, TRPC6 is implicated in the pathogenesis of idiopathic pulmonary hypertension. Three key studies showed that (i) TRPC6 expression was increased in pulmonary artery smooth muscle cells taken from idiopathic pulmonary hypertension patients (Yu et al., 2004); (ii) TRPC6-deficient mice failed to develop pulmonary hypertension in response to chronic hypoxia (Weissmann et al., 2006); and (iii) SNPs in the TRPC6 gene promoter region, which cause elevated expression of the channel, are associated with idiopathic pulmonary arterial hypertension (Yu et al., 2009a).

Skin disorders and TRP channels

TRP channels are present in both neuronal and non-neuronal cells in the skin where they are thought to play a key role in itch, regulation of barrier function, keratinocyte differentiation, hair growth, inflammation and wound healing (see Moran et al., 2011). TRPV1 is expressed in sensory nerves innervating the skin and genetic deletion or pharmacological inhibition of TRPV1 decreased histamine-induced scratching behaviour in mice (Shim et al., 2007; Imamachi et al., 2009). In keeping with this observation, scratching behaviour induced by LTB₄ was decreased by TRPV1 blockade through a mechanism involving attenuated migration of neutrophils to the skin (Fernandes et al., 2013). Moreover, capsaicin injection into the mouse skin pretreated with CFA (but not into healthy skin) induced scratching behaviour, which was attenuated by capsazepine (Liang et al., 2011). There is anecdotal evidence that desensitization to capsaicin creams may reduce itch associated with various aetiologies (Breneman et al., 1992; Ellis et al., 1993; Lysy et al., 2003).

TRPA1 is highly co-expressed with TRPV1 in cutaneous sensory nerves. TRPA1 antagonism and/or genetic deletion, but not TRPV1 blockade, reduced scratching behaviour evoked by intradermal injection of hydrogen peroxide (Liu and Ji, 2012). Chloroquine and BAM8-22 induced itch through a TRPV1-independent mechanism by activating two other receptors, mas-related GPCR A3 (MrgprA3) and MrgprC11. Sensory neurons isolated from TRPA1-deficient mice exhibited markedly decreased responses to chloroquine and BAM8–22 and, accordingly, TRPA1-deficient mice exhibited reduced scratching in response to these pruritogens (Wilson et al., 2011). Moreover, while LTB₄ induced scratching by activating both TRPV1 and TRPA1, the downstream mechanisms leading to itch differ between these channels, i.e. neutrophil migration and superoxide release respectively (Fernandes et al., 2013). These findings suggest that (i) TRPA1 antagonists may be useful for the treatment of itch; and (ii) the spectrum of itch responsive to TRPA1 blockade may be different from that of TRPV1 antagonists.

TRPV3 is abundantly expressed in keratinocytes (Peier et al., 2002b) and is thought to play a key role in barrier formation and hair morphogenesis. Activation of TRPV3 induces release of pro-inflammatory cytokines from murine keratinocytes (Xu et al., 2006a). In human keratinocytes, TRPV3 activation decreased proliferation and induced apoptosis (Borbíró et al., 2011). The same study demonstrated that TRPV3 activation resulted in hair shaft elongation, suppression of proliferation and induction of apoptosis in human organ-cultured hair follicles. Spontaneous mutant rodent strains (DS-Nh mice and WBN/Kob-Ht rats) that carry mutations for Gly573Ser and Gly573Cys in the TRPV3 gene, respectively, display a hairless or a pruritic dermatitis phenotype (Asakawa et al., 2006; Yoshioka et al., 2009). Interestingly, the same mutations together with another for Try692Gly in the TRPV3 gene were identified in patients with Olmsted syndrome, a rare disorder characterized by the combination of peri-orificial keratotic plaques, bilateral palmoplantar keratoderma, alopecia and intense itch. When expressed in heterologous systems, these TRPV3 mutants exhibited gain-of-function phenotypes (Lai-Cheong et al., 2012; Lin et al., 2012). Collectively, these observations suggest that TRPV3 plays a crucial role in skin keratinization, hair growth and itch. Therapeutic intervention by TRPV3 antagonists may reduce keratinization and itch in the skin, and potentially also alleviate alopecia. Of note, we already have a potential tool to test these predictions in GRC15300, a potent and selective TRPV3 antagonist, which is being evaluated in the clinics for pain (Khairatkar-Joshi et al., 2010).

In human keratinocytes, TRPV1 agonist treatment suppressed proliferation and induced apoptosis (Tóth et al., 2011). In human skin, TRPV1 expression was elevated in response to ultraviolet (UV) irradiation (Lee et al., 2009). In the skin of hairless mice, UV irradiation up-regulated the expression of MMPs, pro-inflammatory cytokines, COX and p53; this was reduced by TRPV1 blockade (Lee et al., 2011). Taken together, these findings suggest that TRPV1 antagonists may protect the skin from inflammation induced by UV light such as sunburn. In addition, topical TRPV1 antagonists may ameliorate thermal hyperalgesia after sunburn. TRPV1 has also been implicated in the pathogenesis of atopic dermatitis. Indeed, the TRPV1 antagonist PAC-14028 (AmorePacific Corp., Seoul, South Korea) is being investigated in patients with atopic dermatitis.

Bladder disorders and TRP channels

A number of TRP channels (including TRPV1, V2 and V4, TRPM4 and M8, TRPA1) are expressed in the bladder where they show distinct cellular distribution pattern and play different roles (see Skryma et al., 2011; Avelino et al., 2013).

TRPV1, the target for intravesical vanilloid therapy

TRPV1 is expressed in sensory neurons and urothelium. While TRPV1 in sensory C-fibre afferents is involved in the micturition reflex (Birder et al., 2002) and thought to serve as sensor of painful bladder stimuli (Charrua et al., 2007), the functional role (and the very existence) of TRPV1 in urothelium remains controversial (Everaerts et al., 2010b).

TRPV1-deficient mice fail to develop bladder hyperreflexia during cystitis (Charrua et al., 2007; Wang et al., 2008) and the TRPV1 antagonist GRC6211 attenuates bladder overactivity in a rat model of bladder inflammation induced by LPS (Charrua et al., 2009). GRC6211 also blocked the neurogenic detrusor overactivity (‘neurogenic bladder’) induced by chronic spinalization in the rat (Santos-Silva et al., 2012).

The involvement of capsaicin-sensitive nerves in the human micturition reflex is well-established (Maggi et al., 1989). When the descending neuronal control of the micturition reflex is lost (e.g. after spinal cord injury or due to multiple sclerosis, MS), the bladder becomes autonomic and the capsaicin-sensitive afferents take control of micturition (‘neurogenic bladder’). This forms the foundation for the use of intravesical capsaicin administration in patients with neurogenic bladder: in these patients, capsaicin reduces the first desire to void by increasing bladder capacity and pressure threshold for micturition (Maggi et al., 1989).

Although clinically effective in the long term, intravesical capsaicin is unacceptable for many patients because of the initial burning pain that it causes. Resiniferatoxin is a better-tolerated (much less painful) alternative for intravesical vanilloid therapy. Intravesical resiniferatoxin reduced the number of incontinent episodes (and even restored continence in some) in patients with neurogenic detrusor overactivity of spinal origin (Cruz et al., 1997; Cruz and Dinis, 2007). The beneficial effects of intravesical resiniferatoxin were long-lasting (several months) and reversible. Repeat administration replicated the therapeutic value of the initial treatment. Importantly, biopsies taken from the bladder of patients undergoing intravesical vanilloid therapy did not show any significant histopathological and/or ultrastructural (electron microscopic) alterations.

TRPV4

In the bladder, TRPV4 is abundantly expressed in urothelial cells (Yu et al., 2011). Stimulation of urothelial TRPV4 by stretch and/or hypo-osmolality induces ATP release which, in turn, activates purinergic P2X3 receptors in bladder afferents and evokes the micturition reflex (Birder et al., 2007; Mochizuki et al., 2009; Aizawa et al., 2012). TRPV4 is also expressed in the detrusor muscle. Indeed, GSK1016790A, a potent and selective TRPV4 agonist (Figure 5), induces detrusor muscle contractions even in the absence of urothelium (Thorneloe et al., 2008). Consistent with these findings, intravesical administration of GSK1016790A causes bladder overactivity in wild type, but not in TRPV4-deficient, mice (Thorneloe et al., 2008). Moreover, TRPV4-null mice displayed reduced frequency of voiding, increased urine volume per episode, and spatially altered urine spot pattern (Gevaert et al., 2007). TRPV4-null mice also showed reduced detrusor overactivity in a cyclophosphamide-induced cystitis model. HC-067047, a potent and selective TRPV4 antagonist, reduced micturition frequency in a rat model of cystitis (Everaerts et al., 2010a). Of note, a subpopulation of patients with Charcot–Marie–Tooth disease type 2C, a genetic disease caused by gain-of-function mutation in TRPV4, display bladder urgency and incontinence (Landourè et al., 2010). Collectively, these observations suggest that TRPV4 may represent a useful therapeutic target for the treatment of bladder dysfunction.

TRPM8

A subset of sensory bladder afferents expresses TRPM8 (Shibata et al., 2011) and expression of this channel is elevated after bladder outlet obstruction in rats (Hayashi et al., 2011). In patients with overactive and painful bladder syndromes, TRPM8 is similarly up-regulated in nerve fibres and its expression level significantly correlates with clinical scores (Mukerji et al., 2006a). Intravesical cold saline instillation causes uninhibited detrusor contractions in patients with either idiopathic or neurogenic detrusor overactivity, but not in healthy volunteers (Mukerji et al., 2006b). In rats, intravesical infusion of menthol evokes the micturition reflex (Nomoto et al., 2008). Volume-induced bladder contraction and nociceptive reflex responses to noxious bladder distension are reduced by AMTB, a TRPM8 antagonist (Figure 5; Lashinger et al., 2008). Additionally, in a cystometric study using rats, N-(4-t-butylphenyl)-4-(3-chloropyridin-2-yl)tetrahydropyrazine-1(2H)-carboxamide (BCTC), another TRPM8 antagonist that also blocks TRPV1 and V4 channels, inhibited detrusor overactivity induced by menthol or cold stress (Lei et al., 2013). Taken together, these observations imply that TRPM8 is involved in bladder pain and detrusor overactivity.

TRPA1

TRPA1 is highly expressed in sensory neurons innervating bladder where it is co-expressed with TRPV1 (Streng et al., 2008). As with TRPM8, TRPA1 is up-regulated in bladder mucosa in patients with bladder outlet obstruction (Du et al., 2008). Exposure of rat bladder strips to TRPA1 agonists induced contraction (Andrade et al., 2006). In vivo, TRPA1 agonists increased the micturition frequency in rats, and desensitization of TRPV1-expressing C-fibres by capsaicin attenuated the effect of TRPA1 agonist (Du et al., 2007; Streng et al., 2008). Conversely, the TRPA1 antagonist HC-030031 attenuated bladder overactivity in models of cyclophosphamide-induced cystitis and spinal cord injury (Andrade et al., 2011; Meotti et al., 2013). These findings identify TRPA1 as a potential drug target for bladder disorders.

TRPM4

TRPM4 is functionally expressed in the detrusor muscle of the rat and guinea pig. 9-Phenanthrol, a selective TRPM4 antagonist, reduced contraction of detrusor-isolated strips induced by various stimuli including electrical field stimulation (Smith et al., 2013a,b2013b). Further studies will be needed to elucidate the pathophysiological roles of TRPM4 in human bladder.

Inflammatory bowel disease (IBD) and TRP channels

TRP channels are widely expressed in the digestive tract, with important roles in taste, visceral sensation, GI motility, as well as absorptive and secretory functions (see Boesmans et al., 2011; Holzer, 2011; Blackshaw et al., 2013). Changes in TRP channel expression have been detected in a variety of GI ailments as exemplified by increased TRPV1 expression in both GERD (Matthews et al., 2004; Bhat and Bielefeldt, 2006) and irritable bowel syndrome (Akbar et al., 2008; Keszthelyi et al., 2013). Indeed, TRPV1-null mice develop less oesophagitis after acid exposure compared with their wild-type littermates (Fujino et al., 2006).

There is good evidence to suggest an important role for TRP channels (in particular, TRPV1 and TRPA1) in the development and maintenance of IBD. Increased TRPV1-like immunoreactivity was reported in colonic biopsies taken from patients with IBD, both Crohn's disease and ulcerative colitis (Yiangou et al., 2001). In a rat model of IBD, desensitization to topical capsaicin of intestinal afferents was shown to reduce ulceration (Goso et al., 1993). Furthermore, the TRPV1 antagonist JYL1421 suppressed colorectal distension and improved colitis in rats (Miranda et al., 2007). TRPA1 expression is elevated in the inflamed mouse gut (Yang et al., 2008; Izzo et al., 2012). Experimental colitis induced by dinitrobenzene sulphonic acid (DNBS) was attenuated after both pharmacological blockade (by the TRPA1 antagonist HC-030031) and genetic inactivation of TRPA1 (Engel et al., 2011). DNBS was shown to bind to cysteine residues in the intracytoplasmic N-terminus of the TRPA1 protein, identifying TRPA1 as a direct molecular target in DNBS-induced colitis (Engel et al., 2011). The involvement of TRPA1 in IBD is, however, more complex as TRPA1 activation by cannabichromene (Figure 6) was reported to ameliorate murine colitis (Romano et al., 2013).

In the GI tract, TRPV4 is expressed both in epithelial cells and sensory afferents (Brierley et al., 2008). In IBD patients, increased TRPV4 mRNA levels were reported. In experimental animals, TRPV4 activation contributes to intestinal inflammation via chemokine release and TRPV4 blockade alleviates colitis symptoms (D'Aldebert et al., 2011; Fichna et al., 2012). Finally, TRPM8 is up-regulated both in mouse and human colon during colitis and TRPM8 activation by icilin attenuates inflammatory responses in a mouse model of IBD (Ramachandran et al., 2013).

Diabetes and TRP channels

A growing number of TRP channels (TRPM2, M4 and M5, TRPA1) have been implicated in insulin release from pancreatic beta-cells (see Colsoul et al., 2013). TRPM5 plays an important role in glucose-induced high-frequency oscillations. Indeed, membrane potential and cytosolic calcium level were reduced in islets derived from TRPM5-deficient mice, resulting in decreased glucose-induced insulin secretion (Colsoul et al., 2010). Consequently, TRPM5-deficient mice showed impaired glucose tolerance (Brixel et al., 2010; Colsoul et al., 2010). In addition, TRPM5 mediates fructose-induced insulin release (downstream of sweet taste receptors) in murine islets (Kyriazis et al., 2012). These observations indicate that TRPM5 may serve as a potential convergence point between sweet taste receptors and glucose-induced insulin secretion in pancreatic beta-cells (Kyriazis et al., 2012). In accord with this hypothesis, SNPs in TRPM5 are associated with pre-diabetic phenotypes in subjects at increased risk for type 2 diabetes in a German population (Ketterer et al., 2011). Collectively, these findings suggest that TRPM5 agonists may provide therapeutic benefits in patients with type 2 diabetes.

TRPM2 is another TRP channel implicated in insulin release from pancreatic beta-cells. Pharmacological inhibition or genetic deletion of TRPM2 reduced insulin secretion from islets induced by heat, glucose or glucagon-like peptide 1 (GLP-1) receptor agonists (Togashi et al., 2008; Uchida et al., 2011). TRPM2 is a non-selective Ca²⁺ permeable cation channel and TRPM2-mediated insulin secretion occurs partly by intracellular influx of Ca²⁺. Interestingly, insulin secretion via TRPM2 can be induced by Ca²⁺ influx-independent mechanisms as glucose-induced insulin secretion was lost in islets from TRPM2-deficient mice in a condition that is supposed to completely inactivate the insulin release pathway mediated by K_ATP channel and voltage dependent Ca²⁺ channel (Uchida et al., 2011). Importantly, TRPM2 knockout mice showed higher basal glucose levels and impaired glucose tolerance, indicating TRPM2 agonists may be useful as anti-diabetic agents (Uchida et al., 2011). On the other hand, other studies indicate activation of TRPM2 by ROS leads to apoptosis in beta-cell lines (Hara et al., 2002; Ishii et al., 2006; Lange et al., 2009). In addition, genotyping studies did not find any correlation between SNPs in the TRPM2 gene and type 2 diabetes (Romero et al., 2010). Clearly, further investigation is required in order to establish the validity of TRPM2 as a therapeutic target in diabetes.

The roles of TRP channels in diabetes extend beyond insulin secretion from beta-cells. For example, there is good evidence that TRPV1 in sensory afferents (presumably following the release of sensory neuropeptides) plays a role in physiological glucose control (see Suri and Szallasi, 2008). Furthermore, dysregulated TRPV1 activity was implicated in the pathomechanisms of diabetes, both type 1 (Tsui et al., 2007) and type 2 (Tsui et al., 2011). Capsaicin evokes GLP-1 release from a murine entero-endocrine cell line in vitro and when given intragastrically, capsaicin increases GLP-1 and insulin secretion in wild type, but not in TRPV1-deficient mice (Wang et al., 2012). TRPV1-deficient mice also show improved glycemic control in a diet-induced obesity model (Marshall et al., 2013). Desensitization to capsaicin was reported to increase glucose tolerance in Zucker diabetic fatty rats, a model of type-2 diabetes (Gram et al., 2007). BCTC (a non-selective TRPV1 antagonist that also blocks TRPV4 and TRPM8 channels) improved glucose tolerance in rodent models of type-2 diabetes (Tanaka et al., 2011). Finally, genetic deletion of TRPV1 protected mice from the development of autoimmune (type 1) experimental diabetes (Razavi et al., 2006). Thus, the range of the experimental evidence implies a therapeutic potential for TRPV1 blockade in diabetic patients.

Obesity is an important cause of insulin resistance and type 2 diabetes. The exact role of TRPV1 in the control of appetite and body weight remains controversial. There is anecdotal evidence that dietary capsaicin suppresses appetite and keeps both experimental animals and human volunteers lean. Genetic deletion of TRPV1 appears to protect young mice from high-fat diet-induced obesity (Motter and Ahern, 2008) but when these mice grow older, they become ‘lazy’ (hypoactive) and fat (Garami et al., 2011; Marshall et al., 2013).

The involvement of TRPV4 in body weight control is likewise controversial. While two studies demonstrated that TRPV4-deficient mice were protected from diet-induced obesity (Kusudo et al., 2012; Ye et al., 2012), another study showed the opposite effect (O'Conor et al., 2013). Even worse, in the latter study the TRPV4 (−/−) animals not only became obese but also showed severe, debilitating osteoarthritis.

The potential involvement of TRPM8 in obesity was recently suggested by the finding that chronic dietary menthol treatment prevented diet-induced obesity in wild-type, but not in TRPM8-deficient mice, through thermogenesis in brown adipose tissue, mediated by uncoupling protein 1 (UCP-1), (Ma et al., 2012b).

For the sake of completeness, it should be mentioned here that TRP channels are also potential targets for managing the complications of diabetes, such as peripheral neuropathy, nephropathy, retinopathy, as well as cardiovascular disease. As mentioned earlier, TRPA1 and TRPV1 have been investigated as analgesic targets in diabetic neuropathic pain (Koivisto et al., 2012). Of note, methylglyoxal, a metabolite of glucose, is capable of directly activating TRPA1, providing a potential link between elevated glucose levels and pain (Ohkawara et al., 2012). With regard to diabetic nephropathy, podocyte foot processes and slit diaphragms contribute to the formation of the glomerular filter in the kidney and dysfunction of podocytes leads to proteinuria (Faul et al., 2007). TRPC6 expression is elevated in podocytes by high glucose levels. Moreover, TRPC6 mediates high glucose-induced podocyte apoptosis where TRPC6 is activated by ROS generated from glucose (Liu et al., 2013; Yang et al., 2013). These observations suggest a role for TRPC6 in the development of diabetic nephropathy. Last, TRPV4 in ocular endothelial cells has been implicated in the pathogenesis of diabetic retinopathy (and other ocular neovascularization disorders).

Cancer and TRP channels

The link between TRP channels and cancer is a speculative, potentially rewarding, but highly controversial area of research. Clearly, the expression of several TRP channels (including TRPV1, V2, V6, TRPC3, C5, C6, TRPM1, M2, M3, M4, M7 and TRPM8, TRPA1) is altered in various cancers (see Santoni and Farfariello, 2011; Liberati et al., 2013). Some authors argue that TRP channels are involved in the proliferation and migration of cancer cells, as well as in their resistance to chemotherapeutic agents (see Liberati et al., 2013; Lehen'kyi and Prevarskaya, 2011). Sceptics point out that altered TRP channel expression may be simply an epiphenomenon to cancer progression and not a contributor to disease.

Of note, prostatic adenocarcinoma shows increased TRPM8 expression that appears to correlate with the aggressiveness of the disease. Recently, D-3263, a TRPM8 agonist, has entered Phase 1 clinical trials (NCT00839631) with the hope that it will kill TRM8-expressing cancer cells by calcium and sodium overload through TRPM8 activation (see Santoni and Farfariello, 2011). If this trial is successfully completed, a similar approach may be tried for other cancers that overexpress TRP channels such as TRPC6 in glioblastoma and TRPV2 in ovarian carcinoma.

Kidney disorders and TRP channels

TRPP2, TRPC6 and TRPM6 are causative genes of some genetic kidney disorders (Nilius and Owsianik, 2010). Mutations in PKD2, a gene coding TRPP2, are responsible for ADPKD, the most common inherited renal disease characterized by the growth of numerous cysts in both kidneys, and in many cases, hypertension, leading to renal failure (Peters et al., 1993; Mochizuki et al., 1996). TRPP2 forms a complex with polycystin-1 (Yu et al., 2009b), a gene product of another causative gene of ADPKD, in primary cilia of renal epithelial cells and vascular endothelial cells and this complex is proposed to transduce extracellular shear stress induced by blood flow or urine flow into intracellular Ca²⁺ signals (Nauli et al., 2003; AbouAlaiwi et al., 2009). Cystic renal epithelial cells have lowered intracellular Ca²⁺ level (Yamaguchi et al., 2006) and triptolide, a compound derived from traditional Chinese medicine, suppressed cyst formation by restoring Ca²⁺ signalling in a TRPP2-dependent mechanism in a murine model of ADPKD (Leuenroth et al., 2007).

Interestingly, TRPV4 is reported to form a complex with TRPP2 to serve as a mechano-and thermosensitive molecular sensor in the cilium (Köttgen et al., 2008). Chronic treatment with a low-dose TRPV4 agonist attenuated the renal cyst enlargement in an animal disease model of autosomal recessive polycystic kidney disease, which was also associated with reduced intracellular Ca²⁺ level (Siroky et al., 2006).

Mutations in TRPC6 are responsible for FSGS, a disease characterized by glomerular scarring, resulting in proteinuria, oedema and kidney failure. Some of the TRPC6 mutants cloned from FSGS patients show a gain-of-function phenotype (Reiser et al., 2005; Winn et al., 2005). TRPC6-deficient mice fail to develop angiotensin II (Ang-II)-induced albuminuria (Eckel et al., 2011) while podocyte-specific overexpression of TRPC6 induces kidney dysfunction analogous to human FSGS (Krall et al., 2010). In addition, expression of TRPC6 is elevated both in patients with acquired forms of proteinuric kidney disease and in a model of podocyte injury in rats (Möller et al., 2007). Although the mechanism of TRPC6-mediated podocyte dysfunction has not been clarified, NFAT-dependent gene transcription may be involved in the pathway (Wang et al., 2010). On the other hand, down-regulation of TRPC6 leads to loss of stress fibres upon Ang-II treatment in podocytes (Tian et al., 2010). TRPC6-deficient mice were originally reported to be hypertensive, although compensatory up-regulation of TRPC3 and TRPC7 may have contributed to the observed phenotype (Dietrich et al., 2005).

TRPM6 is permeable to Mg²⁺ and Ca²⁺. TRPM6 is unique among TRP channels in that it has a large carboxyl terminal PK domain, analogous to TRPM7. TRPM6 is primarily expressed in the kidney and intestine and is considered to be responsible for absorption of Mg²⁺. Loss-of-function mutation in TRPM6 is responsible for a rare hereditary disease characterized by profound hypomagnesaemia associated with hypocalcaemia (see Nilius and Owsianik, 2010). Accordingly, TRPM6 agonists may be useful in the management of disorders characterized by hypomagnesaemia.

CNS disorders and TRP channels

A number of TRP channels are expressed in the brain where they are believed to play key roles in the development of neurological and psychiatric disorders (see Vennekens et al., 2012). TRPC3 is a non-selective cation channel that is activated through PLC and activation of inositol trisphosphate receptors. TRPC3 is abundantly expressed in cerebellum, cortex and hippocampus where it plays a pivotal role in brain-derived neurotrophic factor (BDNF)-mediated survival and growth-cone guidance in cerebellar granule neurons (Li et al., 2005; Jia et al., 2007). In addition, TRPC3 can be activated downstream of mGlu₁ receptors and induces slow excitatory postsynaptic potentials in cerebellar Purkinje cells (Hartmann et al., 2008). Interestingly, both TRPC3 knockout mice and ‘moonwalker mice’ (that possess a gain-of-function mutation in TRPC3) exhibit similarly impaired walking behaviours (Hartmann et al., 2008; Becker et al., 2009). Considering this obvious discrepancy, further investigation will be needed to assess whether or not TRPC3 provides a drug discovery opportunity for ataxia.

TRPC6 is a close homologue of TRPC3, and similarly to TRPC3, TRPC6 promotes BDNF-mediated survival and growth-cone turning in cerebellar granular cells (Li et al., 2005; Jia et al., 2007). Overexpression of TRPC6 in cultured hippocampal neurons increases the density of dendritic spine, while down-regulation of TRPC6 with siRNA reduces the spine density (Tai et al., 2008; Zhou et al., 2008). TRPC6 transgenic mice exhibited improved spatial learning and memory in the Morris water maze test, suggesting a crucial role of TRPC6 in learning and memory formation through regulation of synaptic plasticity (Zhou et al., 2008).

TRPC5 is highly expressed in hippocampus and amygdala (Riccio et al., 2009). TRPC5-deficient mice displayed anxiolytic-like phenotype in elevated plus-maze, open field and social interaction tests, but not in novelty-suppressed feeding tests, suggesting that TRPC5 is involved in innate fear (Riccio et al., 2009). In addition, synaptic responses mediated by activation of mGlu receptors and cholecystokinin CCK₂ receptors, both of which are implicated in anxiety, are diminished in lateral nucleus of the amygdala derived from TRPC5-null mice (Riccio et al., 2009). Collectively, these data suggest that TRPC5 may serve as a therapeutic target for anxiety. Parenthetically, TRPV1 (−/−) mice were also reported to show reduced fear and anxiety behaviour (Marsch et al., 2007). This finding, however, has recently been questioned by another study in which only minimal brain TRPV1 expression was found using a sensitive reporter mouse model (Cavanaugh et al., 2011).

Growth cone collapse induced by semaphorin 3A is reduced in hippocampal neurons from TRPC5-deficient mice (Kaczmarek et al., 2012). If this observation is replicated by pharmacological inhibition of TRPC5, it may imply a beneficial effect for TRPC5 antagonists in neurodegenerative disorders.

TRPM2 and TRPM7 are potential therapeutic targets for stroke. TRPM2 is expressed in microglia and neurons in the brain and may serve as a redox sensor. Activation of TRPM2 by oxidative stress could lead to cell death. Indeed, TRPM2-deficient mice showed resistance against neuronal death in a focal ischaemia model of stroke (Miller and Zhang, 2011). Likewise, TRPM7-induced Ca²⁺ overload leads to cell death. In a murine model of brain ischaemia, down-regulation of TRPM7 by shRNA protected rats from ischaemia-induced deficits (Sun et al., 2009). Interestingly, FTY720, an immunosuppressive drug approved for the treatment of MS, inhibits TRPM7 (Qin et al., 2013). It is tempting to speculate that FTY720 suppressed axonal and neuronal loss, one of the main characteristics of MS, partly through its inhibitory effect on TRPM7.

Genetic studies found that SNPs in TRPM2 and TRPM7 genes were associated with two related neurodegenerative disorders, ALS-G and Parkinsonism dementia respectively (Plato et al., 2002). Of note, a different study in a Japanese population did not find significant correlation between SNP in TRPM7 and the ALS-Parkinsonism dementia complex (Hara et al., 2010).

TRPM4 may be involved in axonal and neuronal degeneration in MS. In experimental autoimmune encephalomyelitis, a murine model of MS, TRPM4-deficient mice were protected from axonal and neuronal injury (Schattling et al., 2012). The same study demonstrated that TRPM4-deficient neurons were resistant to excitotoxic stress and energy deficiency in vitro.

TRPML1 is a Ca²⁺-and Fe²⁺-permeable non-selective cation channel expressed predominantly in late endosomes and lysosomes (Nilius and Owsianik, 2010). Loss-of-function mutations in TRPML1 (Dong et al., 2008) are responsible for mucolipidosis type IV, a hereditary lysosomal storage disorder characterized by severe psychomotor retardation, ophthalmologic abnormalities, blood iron deficiency and achlorhydria (Sun et al., 2000; Raychowdhury et al., 2004). The pathological mechanisms arising from mutations in TRPML1 have not been clarified.

Blockade of TRPML1-mediated Ca²⁺ release from late endosome/lysosome vesicles may reduce fusion and trafficking of these organelles (LaPlante et al., 2004). The observed deficiency in lysosome functions could allow accumulation of lipids, followed by impaired autophagosome degradation that normally clears toxic proteins and damaged cell organelles (Vergarajauregui and Puertollano, 2008). Indeed, lysosomal lipid accumulation, defects in membrane trafficking and altered Ca²⁺ homeostasis are common characteristics observed in various lysosomal storage disorders including Nieman-Pick (NP) syndrome. A recent study demonstrated that increasing TRPML1 expression or using ML-SA1, a small-molecule TRPML1 agonist, restored trafficking defects and reduced lysosome storage and cholesterol accumulation in NP type C macrophages, suggesting that TRPML1 agonist may be useful for the treatment of the NP syndrome and potentially, other lysosomal storage disorders (Shen et al., 2012).

Cardiovascular disorders and TRP channels

Cardiac hypertrophy is associated with arrythmias, decompensation and sudden death. Signal transduction pathways that link pathogenic signals to cardiomyocyte hypertrophy may be exploited for therapeutic intervention. There is emerging evidence that the calcineurin/NFAT complex (probably in concert with the MAPK pathway) is one of the key mechanisms that switch on the genes that cause cardiac hypertrophy (see Eder and Molkentin, 2011). Calcineurin is a serine/threonine phosphatase controlled by intracellular Ca²⁺ levels and increases in intracellular Ca²⁺ through TRPC channels (in particular TRPC1, C3 and C6) may activate the calcineurin/NFAT pathway (see Watanabe et al., 2013). Indeed, TRPC1 knock-down by siRNA diminished the hypertrophy phenotype of cultured cardiac myocytes in response to endothelin-1 (ET-1), Ang-II and phenylephrine (Ohba et al., 2007). Consistently, siRNA targeting TRPC1 reduced NFAT activation and hypertrophic response mediated by 5-HT2A receptors in cardiomyoblasts (Vindis et al., 2010). Importantly, TRPC1-deficient mice failed to develop maladaptive cardiac hypertrophy induced by hemodynamic stress and neuro-hormonal excess (Seth et al., 2009). Taken together, these findings imply an important role for TRPC1 channels in the pathogenesis of cardiac hypertrophy.

TRPC3 is another promising target for cardiac hypertrophy. TRPC3 transgenic mice exhibit increased activation of the calcineurin/NFAT pathway, leading to cardiomyopathy and cardiac hypertrophy when challenged by neuroendocrine agonists and/or pressure overload (Nakayama et al., 2006). In keeping with this finding, Pyr3, a selective TRPC3 antagonist (Figure 5), protected mice from pressure overload-induced cardiac hypertrophy (Kiyonaka et al., 2009).

TRPC6 may be also involved in cardiac hypertrophy as knock-down of this gene by siRNAs diminished Ang–II-induced NFAT activation and hypertrophic responses in rat cardiomyocytes (Onohara et al., 2006). Conversely, cardiac-specific overexpression of TRPC6 in transgenic mice led to massive cardiac hypertrophy (Kuwahara et al., 2006).

There is good evidence that TRP channels also contribute to the pathogenesis of hypertension. Expression of TRPC3 is elevated in patients with malignant hypertension in the vascular endothelium of pre-glomerular arterioles (Thilo et al., 2009). ET-1 induces activation of the inositol trisphosphate receptor IP₃R1 in arterial myocytes and causes physical coupling of the IP₃R1 N-terminus to the TRPC3 channel C-terminus, leading to TRPC3 activation and vasoconstriction (Adebiyi et al., 2010). In addition, the kinase WNK4, which is a causative gene of hereditary hypertension, controls blood pressure by restricting TRPC3-mediated Ca²⁺ influx in the vasculature (Park et al., 2011b). Collectively, these observations suggest that TRPC3 blockade may be a novel approach to mitigate hypertension. TRPM4 may be also involved in the control of blood pressure as TRPM4-deficient mice showed a hypertensive phenotype, because of elevated catecholamine secretion from adrenal chromaffin cells (Mathar et al., 2010).

Mutations in TRPM4 are reported to be associated with multiple cardiac conduction disorders, including progressive familial heart block type I (Kruse et al., 2009), isolated cardiac conduction diseases (Liu et al., 2010) and atrioventricular block and right bundle branch block (Stallmeyer et al., 2012). The first two reports demonstrated that mutant TRPM4 underwent reduced deSUMOylation, resulting in constitutive SUMOylation and impaired endocytosis, leading to elevated levsl of TRPM4 channels on the cell surface. Such enhanced surface expression of mutant TRPM4 channel may disturb cardiac conduction by prolonging membrane depolarization and increasing inactivation of Na⁺ channels. On the other hand, heart rates were not altered in TRPM4-null mice while these animals showed hypertension due to enhanced release of catecholamines (Mathar et al., 2010). Lastly, TRPA1 (highly expressed in endothelial cells) has been implicated in the regulation of heart rate and blood pressure (see Earley, 2012).

Conclusions and perspectives

The TRP channel story began in 1969 with the description of a spontaneous Drosophila mutant in which, during prolonged illumination, photoreceptors showed an abnormal, transient response (Cosens and Manning, 1969). Exactly 20 years later, the mutant gene responsible for this abnormal light response was identified and termed ‘trp’ (for transient receptor potential; Montell and Rubin, 1989). 1995 witnessed the discovery of the first mammalian TRP channel, TRPC1 (Wes et al., 1995; Zhu et al., 1995). Within a few years after this seminal discovery, the family of TRP channels exploded to include such long sought-after drug targets as the vanilloid (capsaicin) receptor TRPV1 (Caterina et al., 1997), the camphor receptor TRPV3 (Peier et al., 2002b) and the menthol receptor TRPM8 (McKemy et al., 2002). The initial emphasis of drug discovery efforts was on TRP channels expressed on nociceptive neurons (Patapoutian et al., 2009). Indeed, it took less than a decade to develop the first TRPV1 antagonists to be tried in the clinics as novel analgesic drugs (Szallasi et al., 2007). Antagonists targeting TRPA1 and TRPV3 were quick to follow (Brederson et al, 2013). At the same time, exciting new discoveries have expanded the therapeutic potential of drugs targeting TRP channels into new disease areas, ranging from respiratory diseases (cough, COPD and asthma) through cardiovascular, bladder, metabolic (including obesity and diabetes) and neurological disorders to stroke and cancer.

The rapid progress in TRP channel research has brought the understanding of the roles these channels play in health and disease within reach. However, of the 28 mammalian TRP channels, only 4 (TRPV1 and V3, TRPA1, and TRPM8) have been exploited so far to reach clinical stage of drug development despite accumulating evidence to implicate other TRP channels in diseases. Clearly, several key questions remain to be answered in order to facilitate the translation of the findings in basic research to clinical applications.

First, TRP channels have polymodal gating mechanisms and most also show a broad range of tissue distribution. Even TRPV1 (traditionally considered as a ‘signature of polymodal sensory neurons’) seems to be expressed at unexpected locations such as the skin, urothelium and mast cells (see Szallasi and Blumberg, 1999; Szallasi et al., 2007). Consequently, pharmacological modulation of TRP channels may cause unacceptable, on-target, adverse effects. It was a sobering experience when many TRPV1 antagonists had to be withdrawn from the clinical trials due to either hyperthermia and/or impaired noxious heat sensation (see Moran et al., 2011; Brederson et al., 2013). Of note, TRPV1 antagonists vary significantly in the magnitude of these side effects, raising the possibility that such second-generation antagonists may be synthesized that show an improved clinical benefit to adverse effect ratio.

For other TRP channels, on-target side effects may represent an even bigger problem. For example, blockade of TRPM4 may be beneficial for the treatment of MS (Schattling et al., 2012) and anaphylaxis (Smith and Nilius, 2013), but it may cause dangerous cardiac arrhythmias and hypertension (Abriel et al., 2012). Although both gain-and loss-of-function TRPV4 mutations have been linked to human disease, this channel is another problematic drug target especially when activated by agonists (Nilius and Voets, 2013). Indeed, systemic activation of TRPV4 by GSK1016790A led to endothelial failure and cardiovascular collapse (Willette et al., 2008). To a certain degree, this issue may be circumvented by organ-specific drug delivery: for example, when applied topically to the skin, GSK1016790A promoted intercellular junction development and thus augmented barrier function with no apparent adverse effects (Kida et al., 2012). Selective modulation of TRP channels in diseased, but not in healthy, tissues (e.g. targeting TRPV1-AKAP79 interaction during inflammation) is another attractive approach to circumvent side effects.

Second, most of our understanding regarding the contribution of TRP channels to diseases is derived from studies with in vitro systems or preclinical rodent models, which do not always mirror human diseases. In this context, hereditary diseases caused by mutations in TRP channels (so-called ‘TRP channelopathies’) provide less uncertainty. Unfortunately, gain-and loss-of-function mutations often produce similar phenotypes. In terms of drug discovery, diseases caused by gain-of-function mutations in TRP channels could be more approachable as over-activation of channels could be inhibited by small molecules, while those caused by loss-of-function mutations, particularly truncation types, are difficult to target with small molecules and a less-validated approach such as gene therapy may be required to restore the normal TRP channel function. Besides the direct reversal of dysfunctional TRP channels, modulating the function of the intact TRP channels by therapeutic intervention might provide benefit when diseases related to those channelopathies are caused by mutations in other genes or environmental factors. One such example, although largely speculative, may be the use of a TRPP2 agonist for ADPKD, caused by mutations in PKD1.

Third, for many diseases we already have symptomatic therapeutic modalities and what we really need is a disease-modifying drug. For example, β-agonists improve lung functions in COPD patients, but they do not reverse (or at least halt) disease progression. Likewise, many drugs improve insulin-sensitivity in patients with type 2 diabetes, but these drugs do not fully prevent diabetic complications and neither do they prevent the exhaustion of islet cells.

The advantages and disadvantages of TRP channel agonists and antagonists over currently available therapeutic options need to be carefully weighed. For example, what would be the advantage of TRPA1 and/or TRPM8 antagonists over inhaled glucocorticoids and bronchodilators in patients with asthma? Or how about inhaled TRPV4 antagonists in patients with COPD? One may speculate, but the answers to these questions must come from clinical trials.

These obstacles are real, but probably not insurmountable, and the potential benefits are considerable. Drug discovery companies that can find creative ways to capitalize on targeting TRP channels in disease (and to spare those mediating important physiological functions) may develop novel, first-in-class drugs.

Glossary

ADPKD

autosomal dominant polycystic kidney disease

ALS-G

Guam variant of amyotrophic lateral sclerosis

BDNF

brain-derived neurotrophic factor

CFA

complete Freund's adjuvant

COPD

chronic obstructive pulmonary disease

DNBS

dinitrobenzene sulphonic acid

FSGS

focal segmental glomerulosclerosis

GERD

gastroesophageal reflux disease

GLP-1

glucagon-like peptide-1

IBD

inflammatory bowel disease

WDR

wide dynamic range

Conflict of interest

Yosuke Kaneko is employed by Ono Pharmaceuticals Ltd. The company does not sell any drugs mentioned in this paper.